Help Critique My Cabinet Design, Please!

Greetings!

I am laboring with a design and I just don't know how well I am doing since I am a novice at this.

I am building a 3-way system which has a 15" bass woofer. I have most of the design worked out and cabinets built. Now I am just trying to nail down the woofer selection and crossovers.

I originally designed the cabinets for JBL 15" woofers. However, I am having some questions and I just don't know if I am on the right track.

So, before I blow some serious money I need some help from the experts to review my design and tell me if I am on-track or way off track.

To make it easier for you I created a web site that contains the whole design and how I got to where I am.

I am thinking that the JBL 2235H is the best candidate for this design, but someone else pointed out that the Dayton DC380-8 may be a better fit.

So, I ran both designs in BassBox Pro 6 and XOver Pro 3 to look at both.

Have I done my work correctly? I need your input to tell.

Also, which woofer will be better and why?

Thanks for looking. Here is my design.

I just wanted to say thank you for all of you that took the time to read my appeal for help and to those that actually vised my web site.

This may have not been the best place to ask for those opinions, but nevertheless I appreciate all of your time.

JBL has always been kind of endeared to my heart when I bought my first pair of D-140Fs in high school and put them into my first version of the "Pyramids" long ago. I guess I was looking for some valid rationals to consider using the 2235 over my current woofers and maybe to recapture some of the glory from the old days.

Maybe the only way to really answer that question is to take the plunge and find out for myself. If/when that day happens I will be glad to report my findings back to all of you.

Many thanks,

Loren

I would consider an active crossover between the woofer and mid. A dedicated amp for the 2235 will give you better control over the woofer.

The actual (preferred) tuning can differ from the tuning arrived at with a design program. I would try a few different tunings in both directions of the designed tuning. The cabinet size looks fine for the 2235.

Quote:

---

Originally Posted by yggdrasil

I would consider an active crossover between the woofer and mid. A dedicated amp for the 2235 will give you better control over the woofer.

The actual (preferred) tuning can differ from the tuning arrived at with a design program. I would try a few different tunings in both directions of the designed tuning. The cabinet size looks fine for the 2235.

---

On the money! And, I see you have the signal generator and volt meter so you can run the data and see what's what in the real world; sweep the frequency and record AC voltage across the speaker terminals in convenient increments to generate and plot the actual Helmholtz curve. Tweak the vent diameter and/or tube length to get in the area of 30 - 32 Hz, the typical recommended JBL tuning for the older 15" woofers. Then you can "play" with the tuning, as recommended above, to see what sounds best in your room. Try a closed box too, you may be surprised how good it sounds. JBL used to provide enclosure construction manuals for all this - may be available in the library here on site. Mike

Quote:

---

Originally Posted by Mannermusic

Quote:

---

Originally Posted by yggdrasil

I would consider an active crossover between the woofer and mid. A dedicated amp for the 2235 will give you better control over the woofer.

The actual (preferred) tuning can differ from the tuning arrived at with a design program. I would try a few different tunings in both directions of the designed tuning. The cabinet size looks fine for the 2235.

---

On the money! And, I see you have the signal generator and volt meter so you can run the data and see what's what in the real world; sweep the frequency and record AC voltage across the speaker terminals in convenient increments to generate and plot the actual Helmholtz curve. Tweak the vent diameter and/or tube length to get in the area of 30 - 32 Hz, the typical recommended JBL tuning for the older 15" woofers. Then you can "play" with the tuning, as recommended above, to see what sounds best in your room. Try a closed box too, you may be surprised how good it sounds. JBL used to provide enclosure construction manuals for all this - may be available in the library here on site. Mike

---

Thanks!

Can you elaborate a little on the method to get the frequency response using the AC meter? I understand I can use it to find the actual tuned port frequency because it represents the impedance of the "system" at a given frequency. Is that correct?

How does that relate to the actual response from the system?

I will look up Helmholtz curve and see if that provides some answers, too.

Thanks again. I have so much to learn, but this is really a fascinating little hobby. Did I just say little (not according to my bank account)? ;)

Yes, exactly - impedance is max at max voice coil excursion and reflected by voltage reading. All you do is sweep the frequency and record the voltage at each test point - say every 5-10 Hz. Then plot the resulting curve. It should look similar to the curve generated by your computer model - except it will be correct - no assumptions built in! The "tuning" value is the null between the two humps. Just hook the speaker leads to a sign wave generator/oscillator with a 100 ohm resistor in one lead. Hook your AC volt meter to the speaker terminals. You will need a separate amp if you are using computer signal or other non amplified signal. Your choice.

Technique: 1) turn on the oscillator and set to approx. 200 Hz. 2) Adjust volume to normal listening level (avoid excessive sine wave input to speaker, just normal listening level). 3) Select suitable AC voltage scale on your meter to provide good resolution. Either analog or digital will work but analog is ideal. An old Simpson, or similar. I've done it both ways but the analog enables you to see the voltage rise and fall as you vary the frequency. I typically see a max voltage of 2 - 3 volts, the harder you drive the speaker, the greater the AC signal. Not critical. 4) Take data every 10 Hz or so from 200 Hz down to approx. 10 Hz if your oscillator can handle it. Even the cheap ones can go to 20 Hz. 5) plot your data! Repeat as necessary - as I say, the recommendation for the older woofers was in the area of 32 Hz but you may like a little higher or lower, as yggdrasil suggests above. Depends on your room resonance, etc. Rock and rollers typically like a higher tuning (more BOOM). Personal preference. Ideally, you could generate system response with each porting configuration if you want to optimize system response - that's how the guys in the engineering lab do it. But, truth is, it's how it sounds in your room. And, it's not all that critical, you'll find. JBL used to say that if you are within 10% of optimum you are good enough (can't hear the difference). Bet you end up somewhere between 30-35 Hz. Good luck! Mike

Example of actual test data: 11.4 Hz, low hump and 53 Hz high hump. Null point - 11.4+53 = 64.4 divided by 2 = 32.1 Hz (LE15A in 5 cubic ft. box with 4"D X 7 1/2" long port).

Quote:

---

Originally Posted by Mannermusic

Yes, exactly - ...

---

Excellent! I'll give that a try tomorrow and let you know what I get.

Thanks,

Loren

Alright. I used a program called Fuzzmeasure to generate a slow sine sweep and then plot the response.

The technique is to place a 820 Ohm (that is what I had) resister in series with the speaker wire and then take the signal at the speaker terminals (before the crossover) and run that back into a USB preamp.

Here is the plot:

http://www.mdbq.net/system_Z_033109.jpg

The enclosure should be tuned at 20 Hz and I think the bump at 200 Hz is probably a poor Zobel network?

I think the above plot is accurate because I did the calibration procedure recommended by the software program.

I also captured the phase diagram, but this looks completely wacky to me.

http://www.mdbq.net/system_Phase_033109.jpg

After I did these runs the Dual Pre USB preamp smoked on me. I think there was a short on the PC board since I found a number of solder balls that were never cleaned off the board. Pisses me off because shipping there and back will cost half the price of a new one and I have to wait weeks for a repair. Just venting...

So what does this tell me? I will hazard a guess. It really states the system impeadence across the spectrum as seen by my amp. This is useful stuff since it tells me when I have my zobel network tweaked and it confirms the vent resonance. However, I don't think it represents the actual frequency of the system, does it?

My guess is that I need to consider the efficiancy of the woofer, mid, and tweeter, crossover, and cabinet to get a read on teh actual system plot. Or am I missing something?

Hi Loren -

I'd say you have two challenges - will offer a few "guesstimates" and leave you to your creative juices. 1) I agree the initial plot looks reasonable - what, 65 Hz at the null? Not surprized it is high because, based on my experience, it looks like you have about twice as much port as you need; my next test would be to plug one of the 4" ports and re-run the data. Fine tune from there. 2) Not unusual to spend more time developing the test technique than actually running the data! Suspect your rig may not have sufficient muscle to drive the woofer; the continuous sine wave signal requires a lot of energy. Probably need a power amp of some sort. That's the beauty of using one of the old lab oscillators - they have a lot of muscle built in and the correct output impedance. Heavy duty - probably were very expensive. Mine was an old HP (weighed 23 pounds). Unfortunately, I also just smoked the powersupply last month after decades of use. Will probably have to do something like you are working on for my next mad experiment! Anyway, I think you are on the right track - I'll keep an eye open; want to see what you come up on the rig. The data should be straight forward once you are up and running. I see there's an old HP oscillator for sale on eBay but you might as well get your rig sorted - it will be cool once up and running! Mike

Quote:

---

Originally Posted by Mannermusic

Hi Loren -

I'd say you have two challenges - will offer a few "guesstimates" and leave you to your creative juices. 1) I agree the initial plot looks reasonable - what, 65 Hz at the null? Not surprized it is high because, based on my experience, it looks like you have about twice as much port as you need; my next test would be to plug one of the 4" ports and re-run the data. Fine tune from there. 2) Not unusual to spend more time developing the test technique than actually running the data! Suspect your rig may not have sufficient muscle to drive the woofer; the continuous sine wave signal requires a lot of energy. Probably need a power amp of some sort. That's the beauty of using one of the old lab oscillators - they have a lot of muscle built in and the correct output impedance. Heavy duty - probably were very expensive. Mine was an old HP (weighed 23 pounds). Unfortunately, I also just smoked the powersupply last month after decades of use. Will probably have to do something like you are working on for my next mad experiment! Anyway, I think you are on the right track - I'll keep an eye open; want to see what you come up on the rig. The data should be straight forward once you are up and running. I see there's an old HP oscillator for sale on eBay but you might as well get your rig sorted - it will be cool once up and running! Mike

---

No way I can believe that the vent tuning is 65Hz. That is a 6+ cubic foot cabinet with two vents 3.95" diameter and 10" long each.

I ran the calculations for vent frequency by hand from the formula in the Loudspeaker Design Cookbook and cross checked that with my calculations in BassBox Pro with heavy dampening and both agreed that the vent resonance frequency should be in the lower 20 Hz.

That's a far cry from 65 Hz! In fact, running the calculations manually and in BassBox Pro for a vent tuning of 65 Hz both indicate that it is not possible! The highest vent frequency is about 44.35 Hz for a vent length of 0.02".

So, I think the actual tuned vent frequency really is 20 Hz, but I think the Zobel network is doing something bad at much further up the frequency band at 200 Hz.

Also, the plot was generated by driving the speaker cabinet through my receiver's amp with the tone controls set flat. However, I think the best way is with a signal geneterator and a good AC meter.

Hum-m-m-m. I hear you. Here's one last piece of info for you - actual data this time (enough "guesstimates"). I managed to find the old tuning chart from JBL we used to use and, guess what, for the 136A, which I believe is the home hi fi predecessor to the 2235, for a 5.1 - 6.0 cubic ft. encl, it recommends total port area of 26 sq in with a duct of 8". Sound familiar? So, tends to verify your design and suspicions re the testing rig, Zobel network, etc. Anyway good luck and in the meantime I've gotta get MY oscillator repaired! :blink: Mike

You can read about tuning an enclosure in JBL's manual here on the last few pages ->

http://www.lansingheritage.org/html/...979-manual.htm

http://www.lansingheritage.org/html/.../1970s-kit.htm

Or you can buy this gadget and do it real easy ->

http://www.woofertester.com/wt2product.htm

Thanks guys!!!

I look over those links.

Loren